AI agents can automate many tasks, but fully autonomous systems are not always desirable or safe. In many real-world applications, certain decisions still require human judgment, oversight, or approval. This is where human-in-the-loop (HITL) patterns become important.

Human-in-the-loop systems combine AI automation with human supervision. The agent performs most of the work, but humans step in at critical points to review outputs, provide feedback, or approve actions before they are executed.

In this chapter, we will explore common human-in-the-loop design patterns and how to integrate human oversight into AI agent workflows effectively.